Identification of volatile organic compounds (VOC) and organophosphate flame retardants (OPFR) in building materials

(1)

Identification of volatile organic compounds

(VOC) and organophosphate flame retardants

(OPFR) in building materials

Daniel Duberg

2017-05-29

Exam project, 15 HP, Chemistry

Supervisor: Thanh Wang, Josefin Persson

Examiner: Tuulia Hyötyläinen

(2)

Abstract

Humans today spend most of their time in various indoor settings such as housing, schools and workplaces. The quality of the indoor environment is therefore of great significance for our wellbeing. However, it has been suggested that the indoor environment contains over 6000 organic compounds, such as various volatile organic compounds (VOC). Around 500 of these compounds is believed to be due to emissions from different surrounding building materials such as insulation, plastic film, sealants and flooring. This study targeted building materials from three low energy preschools that were sampled and analyzed for emissions of VOCs and nine different organophosphate flame retardant compounds (OPFR) using a gas chromatograph coupled to a mass spectrometer (GC/MS). Low energy buildings are buildings that is particularly air tight to be so energy efficient as possible. The study uses a qualitative approach and therefore mainly identifies possible contribution from building materials to indoor environment. More than 100 different VOCs was identified and the most noticeable were meta-, ortho- and para-xylene, toluene, n-hexane and propylene glycol, all but the last compound is associated with hazardous health effects. The building materials that emitted the largest amounts of VOCs was sealants and adhesives. Linoleum flooring and acrylic was also large emitters. Tris(1-chloro-2-propyl) phosphate (TCIPP) were identified in all samples and all nine targeted OPFR compounds were identified in the various material samples and dust samples. T-Flex tape and plastic film was the sample materials that emitted most OPFR compounds.

(3)

Introduction

Humans today spend most of their time indoors and the indoor environment have been proven to cause inconvenience, especially among those who are oversensitive due to, for example, allergies and asthma. (1) A house that causes symptoms among its inhabitants is called “sick house” and it is believed that low air exchange rate combined with chemical emissions from building materials is one of the causes to these problems. (2) Since regulations have become stricter for buildings to lower their energy usage, low energy houses have become more common during the last decade due to their energy efficiency and low contribution to the climate. (3, 4) To investigate the potential contribution of chemical emissions from building materials to the indoor environment in low energy houses, three preschools around the Örebro region were selected. From the three preschools, approximately 13 building material samples were collected consisting of flooring, insulation, paint, adhesive, plastic film, sealant and tape. Volatile organic compounds (VOCs) and organophosphate flame retardants (OPFRs) were the chemical compounds of interest in this survey and all materials were analyzed for their

content by two separate methods.

To see possible correlation between VOC emissions from building materials and indoor environment, the material analysis was compared to air samples collected from the same preschools.

Aims of project

- To evaluate VOCs and OPFRs in building materials from three low-energy preschools.

- To evaluate if emissions from building materials contributes to the indoor air quality. - To compare the composition of the identified chemical compounds with those found

in other studies.

Background

Indoor exposure

Humans are currently exposed to a wide range of industrial chemicals in our everyday life and a contributing factor is the amount of time spent in different indoor settings, which is

approximately 20 hours per day. The risk of spending too much time in indoor settings could lead to increased exposure of chemicals that can be emitted from furniture, electronics, textiles and building materials. The high prevalence of industrial chemical compounds indoor is explained by direct emission from building materials, interior and consumer products, lower removal and degradation rates than outdoors due to low biodegradation, low photolysis and consistent indoor climate conditions. (5) The term sick building syndrome is defined as a building that causes various physical discomfort among people that resides in the building, and the discomforts include headache, drowsiness, respiratory problems, and more. The reasons causing these problems could generally be associated with structural dampness and low air circulation, often in combination with the emittance of chemical compounds from building materials and/or furniture. (2) The term and symptoms are, however, rather blurry as the perception of symptoms have proved to be affected by various factors such as

psychological, gender and occupational conditions. (6)

During the last century, the sources of chemical compounds and exposure have shifted considerably. With the same pace as chemicals of concern are being phased out due to stricter legislations, new chemicals are introduced to replace the old ones, often without sufficient risk assessment. However, the replacement chemicals are often not sufficiently tested in regard to their environmental health and toxicity influences and thus can be found to have adverse effects long after their introduction and usage. (7)

(6)

There are different exposure pathways of industrial chemical compounds in indoor environments such as breathing, direct skin contact of dust or materials or inadvertent intake of dust through the digestive system (8). Children are at greater risk of exposure to harmful chemicals than adults since they eat, drink and breathe more than adults relative to their size, they also have more direct contact with toys, clothes and electronic devices that can contain significant amounts of hazardous chemicals. Many toxic organic compounds are also

semivolatile and therefore tend to partition to dust due to their lipophilic properties. This can pose a greater risk to toddlers that crawls around on floors and have explicit hand-to-mouth behaviors, and other exposure pathways may also include breastmilk since many compounds have been proved to be bioavailable (7, 8, 22). Children are especially vulnerable to toxic chemicals in our environment since some of these compounds, so called endocrine disruptors (EDs), can interfere with normal hormone function which is crucial for the development. It has also been found that some chemicals can cross the placenta and be transferred to the fetus from the mother (7). In an attempt to decrease the exposure of toxic chemical to children, the Swedish Chemicals Agency (Kemikalieinspektionen) has developed an action plan especially targeting preschools as a priority area. The action plan was ordered by the Swedish

government and is called Handlingsplan för en giftfri vardag 2011-201- skydda barnen bättre. (9)

Indoor air quality

To determine the indoor air quality (IAQ), several factors should be taken into account, such as the emission of chemical compounds, dampness, temperature and radon to obtain a full evaluation of the buildings status. There is a large variety of sources of indoor chemical emissions and these include type of building material, cleaning activities, furnishing, use of cosmetics, cooking, household products, and more. New chemical compounds may also be formed as a result of chemical reactions between different pollutants in the environment, and in damp environments microbial processes may also occur which can affect the indoor air quality (1, 2). The largest group of these emitting chemical compounds is classified as VOC and approximately one thousand VOC have been identified in indoor environments. (2) Flame retardants can also contribute to the contamination of the indoor environment and have been reported to be present in both air and dust since many flame retardants are semi volatile organic compounds (SVOCs). (10, 11)

Low-energy buildings

Buildings are responsible for a large amount of the energy consumption in the modern society. In Sweden, they stand for approximately 30% of the total energy consumption. This have resulted in directives issued by the European Union (EU) which states that the total energy consumption in buildings needs to be reduced and the total energy must be used in a more effective way. Therefore, these new directives state that all new buildings must be so called nearly energy buildings by the end of 2020. (3, 4) The requirements for zero-energy buildings is, in short, to reduce zero-energy consumption to a minimum and contribute to the reduction of carbon dioxide released to the environment. To be able to meet these requirements the buildings are in general particularly airtight and efficiently insulated to prevent heat from escaping from the building which result in decreasing energy consumption. (12)

Building materials

Over 6000 different organic compounds have been identified in indoor environment and about 500 of these compounds is believed to be emitted by various building products. (8) The building materials are divided into different classes dependent of their properties, for example framework materials, insulation materials, coating materials and protective materials. This

(7)

study focused on the materials that are associated with indoor environment, such as

insulation-, coating- and protective materials, e.g. insulation, plastic film, paint, sealants and flooring. (2)

Flooring is often a mixture of various components rather than one single material, for example, PVC, linoleum, adhesives, textiles, primers and these have been shown to emit VOCs, even though they tend to wear off during a near future after montage. Examples of chemical compounds identified are phenols, alcohols, alkanes, aromatic compounds, phthalates and benzenes, such as 2-(2-butoxyetoxy) ethanol, butoxyethanol, hepta- and octadecane and tributyl benzene, and 2-ehtylhexanol. (1, 8, 13)

Paint often contain different organic solvents such as n-butyl acetate, 2-(2-butoxyetoxy) ethanol and propylene glycol and may appear in indoor air for a long extend of time after application. Other noticeable compounds include alkanes, glycol esters and glycols such as propanal, butanal, pentanal, hexanal and texanol. (1, 8, 13)

Adhesives have been shown to emit alkanes around C9 to C11, toluene and styrene. Compounds found in sealants include ketones, esters, glycols, siloxanes, styrenes and polychlorinated biphenyls (PCBs). (8, 13)

Emission testing

To determine potential chemical emissions from materials to the indoor environment, proper equipment is crucial to maintain stable physical conditions to obtain realistic emission testing values. Main parameters that are taken into consideration are temperature, air exchange and humidity. Climate controlled emission testing systems have been developed to meet these requirements and can be divided into two techniques, emission test chamber and emission test cell. The emission test chamber is a sealed container, often stationary, at various sizes that ensure isolation and minimal external influence. The emission test cell is a portable device which can be mounted directly to the surface of the material of interest. (14)

Emission rate is used to calculate the final concentration of emission from samples, and is determined by measuring the concentration of emission from the material over time, (µg/m2, h). This can be done by placing a material in an emission test chamber and measure the concentrations at fixed intervals. (15, 32)

Volatile organic compounds (VOCs) and semi-volatile organic compound (SVOC) The definition of VOCs is in this study referred to the standards prEN 16516 and ISO 16000-6, and states that an organic compound that has a boiling point between 50°C to 260°C measured at a standard pressure of 101,3 kPa is considered as a VOC. It also states that compounds that elute between n-hexane and n-hexadecane in the gas chromatograph column (GC column) specified in prEN16516 are also included as VOCs.

SVOC is also defined according to the standards prEN 16516 and ISO 16000-6, and states that organic compounds with boiling points ranging from 240°C to 400°C is considered SVOCs if they elute after n-hexadecane in the GC column specified in prEN16516. (16, 17) Due to their low volatility, most OPFRs are classified as SVOCs. (18) Sources of VOCs in indoor environments include outdoor air, building materials, furnishing, smoking and various consumer products. (19)

OPFRs

Organic flame retardants (FRs) have been widely used in furniture, electronics, building materials and other products, and they have been more extensively used during the last decades. Even though it has been suggested that many flame retarding compounds pose a threat to human health, their negative effects are often overlooked because their flame retardant properties were considered to be more advantageous in preventing fire associated

(8)

injuries and deaths than possible health risks associated with their chemical properties. (20) Brominated flame retardant (BFRs) compounds are characterized by its multiple bromine atoms attached to a carbon structure and they are used as additives to various products to improve fire resistance. The electronegative bromine is attracting free radicals that are

released during combustion of organic materials, thereby reducing the spreading of fire. Since the BFRs are additives, they are not strongly bound to the materials and therefore tend to leach out to the surrounding environment. (21) One class of BFRs is polybrominated diphenyl ethers (PBDEs) which can contain one to ten bromine atoms giving a total of 209 possible congeners. PBDEs have however been classified as persistent organic pollutants (POP) due to their persistency, bioaccumulative and toxic properties, and have been listed under Annex A of the UNEP Stockholm convention which bans further usage and production of the

compounds. To maintain the fire safety regulations for products on the market, PBDEs have been replaced by other compounds with flame retarding properties. Organophosphate flame retardants (OPFRs or PFRs) have increased in usage as flame retardants because of their broad chemical properties such as their compatibility with many polymers and other chemical compounds. (22) OPFRs such as the non-chlorinated alkyl phosphates are primarily used as plasticizers, while chlorinated alkyl organophosphates are mainly used as a substitute for pentaBDE formulations and aryl phosphates are also used as flame retardants (23).

The knowledge about the influence of OPFRs on human health is today rather deficient and few studies have been conducted in the matter although numerous of studies on laboratory animals have reported adverse effects such as carcinogenic effects and disturbances in reproductivity. (20, 22) Even though OPFRs show lower persistency and bioaccumulativity than the BFR compounds some OPFRs have been classified as persistent in aquatic

environments and have even been detected at higher levels in indoor and outdoor

environments in matrixes such as water, air, soil and sediments. (22, 23) The following OPFR compounds have been associated with health risks such as carcinogenic, weakening sperm quality, osteogenic disruption among others; tris(1,3-dichloro-2-propyl) phosphate (TDCIPP), tris(2-chloroethyl) phosphate (TCEP), triphenyl phosphate (TPHP) and tris(2-butoxyethyl) phosphate (TBOEP) (24) This study will also analyze triethyl phosphate (TEP), 2-ethylhexyl diphenyl phosphate (EHDPP), tris(1,3-dichloropropyl) phosphate (TDCPP) and tris(propyl) phosphate (TnPP). (25, 26)

As mentioned, OPFRs are currently widely used as flame retardants; in 2006 they accounted for 11.5% of the world consumption of flame retardants which is about 200 000 tons, and in the EU they accounted for 20%. (10)

(9)

Materials and methods

Selected materials

The materials of interest were collected from three different preschools located in Sweden which were built according to three different building techniques, see Table 1.

Table 1 |List of the three low energy preschools and the materials taken from each preschool. Low-energy preschool A Low-energy preschool B Low-energy preschool C Building year 2015 2015 2016

Building type Two story house Two story house Two story house

Area (m2₎ ₁₀₃₂ ₁₀₄₄ _695.5

Building technique Small energy house Passive house Small energy house

Certification Non Swan labeled Environmental building silver

Materials T-Flex tape, Plastic film Insulation, Paint, 3M tape, Plastic film, Linoleum flooring Sealant, Acrylic, Insulation, Plastic film, Linoleum flooring

Different building materials such as flooring, plastic tape, sealants and wall paints were selected for analysis to obtain an overall representative picture of the surrounding

environment of the preschool rooms. Some of the materials, such as plastic film, linoleum flooring and paint, constitutes large areas of the construction and are therefore of interest. The stickier materials such as, sealant and tape, are interesting for this analysis because they are made of various organic compounds and might therefore be a source of emitted compounds. The low amount of material samples from preschool A is due to the fact that there were almost no materials left to collect at the date of sampling because the carpenters have cleared the area after the construction were finished.

Materials OPFRs

Separate mixtures of native and isotope labelled OPFRs; tris(chloroethyl) phosphate (TCEP), tris(2-chloroisopropyl) phosphate (TCIPP), tris(1,3-dichloroisopropyl) phosphate (TDCIPP), triethyl phosphate (TEP), tri-n-propyl phosphate (TnPP), tri-n-butyl phosphate (TnBP) were from Cambridge Isotope Laboratories, Inc. (Tewksbury, MA, USA), while 2-ethylhexyl diphenyl phosphate (EHDP) and tris(2-butoxyethyl) phosphate (TBOEP) from Chiron AS (Trondheim, Norway) was used. The recovery standards 13C-triphenyl phosphate (MTTP) and

13_{C-PBDE-77 and}13_{C-PBDE-138 were purchased from Wellington Laboratories (Guelph,}

ON, Canada). Solvents used in this method were hexane (Suprasolv) from Merck KGaA (Darmstedt, Germany), acetone (p.a.), ethyl acetate (chromosolv) and iso-octane

(chromosolv) were from Sigma Aldrich (Schnelldorf, Germany), dichloromethane (Honeywell). SPE columns used in fractionation were Supelclean ENVI-Florisil

(10)

Sigma Aldrich and the sulphuric acid was from VWR International, which were washed with hexane to ensure that no contamination occurred during sample pretreatment. Standard reference material (SRM 2585) was purchased from National Institute of Standards and Technology (Gaithersburg, Maryland, USA) and were used to evaluate the method. Novel brominated flame retardants (nBFRs) were also used in the sample preparation method but these have not been analyzed and thus not included in this study.

VOCs

The internal standard for VOC analysis was trimethyl pyridine (Eco Scientific), and all compounds were calibrated against toluene (Eco Scientific) yielding levels in the unit μg/m3

toluene equivalents. The calibration standard also contained n-decane (Fisher Scientific), 1,3,5-trimethylbenzene (Sigma-Aldrich) and 2-ethyl-1-hexanol (Fisher Scientific). Emission sampling

For the emission sampling, glass flasks about the size of 150 cm3_{were used. To seal the flask}

20 mm butyl injection stopper (Grey) from Scantec Nordic (GBG, Sweden) or 20 mm rubber stoppers (Red) from Alltech Associates inc. (USA) was used and 20 mm Standard metal seals from Alltech Associates inc. (USA) was used to seal the flask.

ORBO-609 XAD-2 (25/50) 400/200 mg from Sigma Aldrich (Schnelldorf, Germany) was used for absorption of OPFRs and Tenax TA sorbents (poly (2, 6-diphenyl-p-phenyl oxide)) from Markes International, UK was used for VOCs.

Method

Emission tests of building materials

The emission test for VOCs was performed using emission test chambers which consisted of glass flasks with the size of 150 cm3. The building materials were cutinto smaller pieces so that they would fit into the glass flasks before they were air-sealed by a lid. The samples were then let to stand for approximately two days before sampling began. The sampling was

performed by pumping helium, since it is an inert gas, through a syringe into the glass flask at a flow of 100 mL/min, the exit flow of the flask went through a different syringe and by that passing a Tenax TA during 15 min.

The emission sampling for OPFRs was performed in the same way as VOC except that ORBO- 609 XAD-2 was used instead of Tenax TA and a flow rate of 250 mL/min was used. The sampling was performed during 15 min for the first samples but was increased to 30 min to make sure that enough air volume was sampled through the absorbent tube. Some OPFR sampling were performed using metal foil bags that were welded airtight, instead of the glass flask, since some materials were stickier than others which made it difficult to insert these into the flask without adhering to the flasks inner walls. The procedure was otherwise the same as for the flasks, see figure 1 and 2.

(11)

Figure 1 | Emission sampling of building materials using the flask method.

Figure 2 | Emission sampling using the foil bag method.

Sampling and analytical procedure of OPFRs

To extract the OPFRs from the sorbent, an extraction method based on Cequier et al.(23) was performed with each batch consisting of four samples, one SRM and one procedural blank. After the emission sampling, the sorbent was transferred to a 15 mL Falcon tube where it was extracted with 4 mL hexane/acetone (3:1; v/v) and internal standards (25:25:25 µL (nBFR AIS, nBFR BIS and OPFR IS)) were added, and the samples were then placed on a

(12)

shaker for 1 hour. The samples were then sonicated for 10 minutes and centrifuged at 3200 rpm for 2 minutes. The supernatant was collected in a 20 mL glass vial after each extraction performed 3 times yielding approximately 12 mL of supernatant. The samples were

evaporated to dryness under a stream of nitrogen gas and 1 mL of hexane was added. To separate the analytes of interest, fractionation using Florisil (500mg/3mL) was conducted to yield two fractions. The columns were washed using 6 mL of hexane before the sample was added and the first fraction, containing the non-polar nBFRs, was eluted using 8 mL of

hexane. The fraction was then placed to evaporate down to approximately 1 mL for further purification. The second fraction, containing OPFRs, was eluted by adding 10 mL of ethyl acetate.

A SPE column filled with approximately 600 mg of silica/sulphuric acid 40% (w/v) was washed with 8 mL hexane before fraction one was placed in the column and the elution was performed using 10 mL of hexane/dichloromethane (1:1; v/v). Both fraction one and fraction two was evaporated down to dryness before they were reconstituted with 100 µL iso-octane and 25 µL recovery standard (MTTP 1 ppm) and moved to GC-vials for analysis.

The analysis was performed with an Agilent 7890A gas chromatograph (GC) coupled to an Agilent 5975C mass spectrometer (MS) equipped with a DB-5ms column (30 m x 250 mm x 0.25 µm) from Agilent technologies, by injecting 1 µL of each sample in splitless mode with a flow of the carrier gas (He) at 1 mL/min. The temperature program had the injection temperature of 90 ˚C and hold for 2 min, followed by a temperature ramp of 10 ˚C/min to a temperature max of 310 ˚C for a final hold of 8 min. The ion source was in electron impact (EI) mode with source temperature at 230 ˚C and the electron energy set at 70 eV. The quadrupole and auxiliary temperature was set to 150 ˚C and 280 ˚C, respectively. Solvent delay was set to 4 min and the MS was operating in single ion monitoring (SIM) mode with a dwell time of 30 ms.

Analytical procedure of VOCs

A gas chromatograph (GC, 7890B, Agilent Technologies) and mass spectrometer (MS, 5977A, Agilent Technologies) coupled with an automated thermal desorber (ATD, TD-100, Markes International, UK) was used and the GC was equipped with a DB5-ms UI column (60 m x 0.250 mm x 1.00 µm) from Agilent Technologies. The ATD was set to 150 ˚C for

desorption of the compounds from the Tenax TA sorbent to avoid any breakdown of the compounds. The samples were injected in splitless mode and the carrier gas (He) had a flow rate of 2 mL/min. The compounds were ionized with electron ionization (EI) set to 70 eV and a source temperature of 230 ˚C. The GC run with the following temperature program; 50 ˚C and hold for 2 min, ramp to 170 ˚C at 5 ˚C/min then ramp to 290 ˚C at 30 ˚C/min and a final hold for 12 min. The transfer line temperature was set to 250 ˚C and detection was performed in full scan mode over the m/z range 29-400.

Dust and air sampling, preparation and analysis

The sampling, preparation and analysis of air and dust from the selected preschools was performed by Josefin Persson (MTM, Örebro University). The sampling, sample preparation and analysis of VOCs followed the international standard protocols ISO 16000-6:2011 (16) and ISO 16017-2:2003 (27). Passive sampling of air was collected over a two-week period and later analyze for VOCs. For OPFRs, the sample preparation and analysis was performed according to the methods presented by Cequier et al. (23) and Van den Eede et al. (30) Dust samples were collected from each preschool during two occasions, one at the beginning of the sampling campaign and the second sample was collected after two weeks. Both the dust and air sampling was conducted four times a year, one for each season.

(13)

Results

A brief summary of selected identified VOC and OPFR compounds from the emission tests of the materials are displayed in Tables 3-5 and Table 6-8. Since a large number of different compounds were identified, they were divided into larger chemical groups as can be seen in Figure 3. However, a few selected compounds are further discussed in the discussion section. The results for both OPFR and VOC emissions are qualitatively and semi-quantitatively presented (see discussion for explanation). Figure 4 displays the total volatile organic compounds (TVOC) from each sample obtained from the GC/MS analysis and shows which materials emitted the largest amount of VOCs compared to the others.

In this study, the VOCs emitted from the building material samples were compared to air samples collected in the preschools since VOCs are volatile and tend to reside in the air, while the OPFRs from the emission tests were compared to dust samples collected from the

preschools since most OPFRs are semi volatile and tend to reside in dust. The air and dust samples were sampled and analyzed by Josefin Persson for her PhD work and the results were not reported in detail in this study.

Figure 3 | The graph shows a summary of the total amount of collected chemical compounds divided into larger chemical groups. The chart shows the portion of all identified compounds from all material samples.

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Distribution of various VOC groups

Glycol ether, Alcohol Alcohol Alkene Ketone

Alkane Aldehyde Ester Siloxane

Carboxylic acid Aromatic hydrocarbon Amine Silane

(14)

Figure 4 | Total volatile organic compounds (TVOC), the chart shows the contribution of each sample according to TVOC. (A) = Preschool A, (B) = Preschool B, (C) = Preschool C

Emission of VOCs

The GC/MS analysis of VOC indicated that the dominating compounds in the building materials analyzed were alcohol, silanes, siloxanes, alkenes, alkanes and aldehydes. Amines, esters, carboxylic acids, ketones and various unidentified compounds were also detected but in much smaller quantities. The dominating emitting materials were sealant and adhesive (Figure 4). The most noticeable chemical compounds according to peak area size was

propylene glycol, trimethoxyphenolsilanes, 1-(2-butoxyethoxy) ethanol, toluene and hexanal, see table 9, in appendix A. Chromatogram of selected materials analyzed can be seen in figure 8-10 in Appendix B.

TVOC

T-Flex Tape (A) Plastic Film (A) Insulation (B) 3M Tape (B) Linoleum Flooring (B) Plastic Film (B) Paint (B) Adhesive (B) Sealant (C) Linoleum Flooring (C) Plastic Film (C) Acrylic (C) Insulation (C)

(15)

Figure 5 | The chart shows the proportion of the identified VOCs in each sampled material.

0% 20% 40% 60% 80% 100%

Plastic film T-Flex Tape

Preschool A

Phenols Various 0% 20% 40% 60% 80% 100%

Insulation Paint Adhesive Plastic film 3M Tape Linoleum Flooring

Preschool B

(16)

Alkanes were highly present in the plastic film samples from all three preschools that

otherwise showed distinct differences in content, especially the plastic film from preschool A. Glycol ether were present in the plastic film from preschool A and B but were not present at all in the preschool C sample. The two tapes from preschool A and B showed similar portions of chemical compounds and the dominating groups in both samples were alcohol and ester. The insulation, however, from preschool B and C showed distinct differences in content. The preschool B sample indicated high presence of alkanes and amines while the preschool C sample contained alkanes, aromatic hydrocarbons and various unidentified compounds. Sealant, adhesive and acrylic are three similar materials, the sealant and acrylic contained similar chemical compounds such as high presence of glycol ethers and alcohols, while the adhesive contained alkanes, siloxanes and various other compounds. The dominating

chemical in linoleum flooring were aldehydes but otherwise they showed some differences in content with the sample from preschool C showed high presence of ketones while it was not at all as dominating in the sample from preschool C. The paint contained large proportions of glycol ether and alkenes.

Table 2 | The table shows the threshold settings, number of peaks identified and TVOC for each material sample and for the two blanks used.

Sample

Preschool Threshold Peaks TVOC (total peak area)

T-Flex tape A 20 15 1,55*109 Plastic film A 19 19 5,13*108 Insulation B 20 11 1,47*109 3M tape B 21 27 4,19*109 Linoleum flooring B 21 50 1,14*1010 Plastic film B 19 30 7,06*108 Adhesive B 22 74 7,26*1010 Paint B 20 29 1,6*109 0% 20% 40% 60% 80% 100%

Sealant Insulation Acrylic Plastic film Linoleum flooring

Preschool C

(17)

Sealant C 22 52 7,45*1010

Linoleum flooring C 21 60 1,01*1010

Plastic film C 19 18 2,60*108

Acrylic C 20 34 1,13*1010

Insulation C 20 17 1,62*109

Blank red cap - 20 3 1,22*108

Blank grey cap - 20 10 1,30*109

Table 2 shows the threshold settings used for each compound, the number of peaks is also interesting since it shows the number of peaks identified in each material sample, one peak represents one chemical compound. The insulation in preschool B for example shows as little as 11 peaks while the adhesive from the same preschool showed the largest number, with 74 peaks. The TVOC for each sample was also interesting since it shows the differences in total peak area between the different material samples and that it may differ in large values. For example, the TVOC from the plastic film from preschool C were approximately a hundred times smaller than the TVOC from the acrylic from preschool C. The threshold settings for the material samples were selected individually for each sample so that the peaks were above the background noise.

Emissions of OPFRs

The compounds identified in the analysis for OPFRs are shown in Tables 3-5.

Because more than two blanks were available, their mean value (𝑥̅) and the standard deviation (σ) was calculated to be able to calculate the method detection limit (MDL) which is three times the standard deviation. The result is valid and can be reported when:

(𝑠𝑎𝑚𝑝𝑙𝑒 𝑣𝑎𝑙𝑢𝑒 −

𝑥

̅ )

> (3×𝜎) Equation 1

The mean value and standard deviation were calculated from the four procedure blanks, one for each batch, and then MDLs were calculated according to Table 10 in Appendix C. The procedure blanks in the analysis of OPFR compounds showed high contamination of the various OPFR compounds.

To calculate the recovery percentage of the internal standard, the following calculation was performed:

𝐶𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑡𝑒𝑐𝑡𝑒𝑑

𝐼𝑛𝑖𝑡𝑖𝑎𝑙 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛 ×100 Equation 2

Table 3 | The table shows the amount of compounds (pg) identified in the analysis. Compounds that did not exceed the MDL values is marked as <MDL, the table also shows if the compound of interest was detected in the dust samples.

Preschool A

T-Flex tape Plastic film Dust

TEP - - Yes TnPP - - No TnBP 14404 <MDL Yes TCEP 870 73 Yes TClPP 2648 3563 Yes TDClPP <MDL <MDL Yes

(18)

TPHP 826 <MDL Yes

TBOEP 873 1215 Yes

EHDPP 727 <MDL Yes

Table 4 | The table shows the amount of compounds (pg) identified in the analysis. Compounds that did not exceed the MDL values is marked as <MDL, the table also shows if the compound of interest was detected in the dust samples.

Preschool B

Insulation Paint 3M tape Plastic film Linoleum flooring Dust

TEP - - - Yes TnPP - - - Yes TnBP 3463 <MDL <MDL <MDL <MDL Yes TCEP <MDL 223 261 251 73 Yes TClPP 3874 4308 2951 2196 1984 Yes TDClPP <MDL <MDL <MDL <MDL <MDL Yes TPHP <MDL <MDL <MDL <MDL <MDL Yes TBOEP <MDL 668 1346 2054 1226 Yes EHDPP <MDL <MDL <MDL <MDL <MDL Yes

Table 5 | The table shows the amount of compounds (pg) identified in the analysis. Compounds that did not exceed the MDL values is marked as <MDL, the table also shows if the compound of interest was detected in the dust samples. *The plastic film sample also contains a blank as a result of human error during fractionation. See discussion.

Preschool C

Plastic film* Linoleum flooring Acrylic Sealant Dust

TEP - - - - Yes TnPP - - - - Yes TnBP 2359 3134 <MDL <MDL Yes TCEP 603 69 <MDL 713 Yes TClPP 17738 2882 3112 2217 No TDClPP <MDL <MDL <MDL <MDL Yes TPHP 846 <MDL <MDL 838 Yes TBOEP 1868 <MDL <MDL 593 Yes EHDPP 3068 <MDL <MDL <MDL Yes

The compounds identified in the analysis for OPFRs are shown in Tables 3-5. The

compounds TEP and TnPP did not show any trace in the analysis and the recovery for the two compounds were very low, 0% to 42%, and TDCIPP did not show presence in any of the material samples. The only compound that was identified in all the provided samples was TCIPP. Even though some of the other compounds were identified in some samples the overall presence of OPFR compounds was low. The sample that contained the most identified OPFRs was T-Flex tape and plastic film from preschool C, with six identified compounds and the sample with the least OPFRs was the acrylic with one identified. The sealant, however, emitted four groups, TCEP, TCIPP, TPHP and TBOEP. The plastic film samples from preschool A and B showed similar content with TCEP, TCIPP and TBOEP identified, while the plastic film sample in preschool C showed content of TnBP, TCEP, TCIPP, TPHP, TBOEP and EHDPP. Differences in emitted OPFRs in T-Flex tape and 3M tape were

(19)

conspicuous as 3M tape did just emit TCEP, TCIPP and TBOEP while T-flex tape did emit TnBP, TCEP, TCIPP, TPHP, TBOEP and EHDPP. The two samples of linoleum flooring emitted different OPFRs, the sample from preschool B emitted TCEP, TCIPP and TBOEP while the preschool C sample emitted TnBP, TCEP and TCIPP. Paint and insulation were low emitters with just three and two emitted compound groups respectively.

Almost all OPFRs were detected in the dust samples collected from the preschools, TnPP was not detected in preschool A and TCIPP was not detected in preschool C whereas all OPFR compounds was detected in preschool B. (except TEP and TnPP)

The recovery of internal standards showed a large variety in values stretching from 60% to 134 % for TnBP, TCEP, TClPP, TDClPP and TPHP and the recovery for TEP and TnPP showed values stretching from 0% to 42%, however, the values were often 0%.

Emissions from building materials according to the literature

Emissions of compounds detected in the analysis performed in this study were compared to compounds found in other studies or reports,Tables 6-8.

Table 6 | Compounds identified in this study and in at least one other study or reference.

Pe = Pegasus lab, VTT = Järnström, H.13_{, SS = Socialstyrelsen}1_{, IP =kemikalieinspektionen}8_{, FS = Identified in preschool air}

samples.

Tabell 7 | | Compounds identified in this study and in at least one other study or reference.

samples.

Preschool A Reference Literature Hazardous

1-butanol PE, VTT, FS, SS X 2-Etylhexanol VTT, X 2-Ethylhexylacrylate PE Hexanal PE, SS Tetradecane FS X Tridecane PE, FS X

Preschool B Reference Literature Hazardous

1-Butanol FS, SS, VTT X 2,2,4,6,6-Pentamehtyl heptane PE X Benzaldehyde FS X Decanal PE, VTT, FS Dodecane VTT X m-Xylene FS, SS X n-Hexane SS, IP, FS X Nonanal PE, VTT, FS X o-Xylene FS, SS X Propylene glycol FS p-Xylene FS, SS X Tetradecane VTT X Toluene SS, IP, FS X Tridecane PE X

(20)

Tabell 8 | | Compounds identified in this study and in at least one other study or reference.

samples.

The reports and studies used for reference have identified various chemical compounds that were not identified in this study, some compounds, however, have been identified in one or more studies. Sixteen of the compounds that have been identified in this study and other studies were also detected in the indoor air samples from the preschools. Three VOCs were detected in materials from all three preschools, 1-butanol, tetradecane and tridecane. Those three compounds were also detected in some air samples from the preschools.

Discussion

Emissions of VOCs

The quantification of VOC levels was not performed in this study due to several of reasons. First of all, there was only a limited number and amount of material available from the preschools which made it difficult to use the same amount of mass in the emission test. There were also some difficulties to control temperature, dampness and airflow due to the basic design of the emission test chamber. However, the lack of emission-rates for each material provided the biggest obstacle to quantify concentrations of VOCs in the samples, since there is no knowledge about at what rate and concentrations that the specified material emits. The emission from materials was also not only dependent on the amount of sample but also the specific surface area that was tested, so rather than compare materials according to weight it would be more accurate to analyze the emission based on the surface area. The reason for this was that a sample with the same amount of material but spread over different area sizes will emit various amounts of compounds. A larger sample area has a greater ability to emit larger quantities of compounds then a smaller area.

Even though it is not possible to accurately determine the concentration of different VOCs, it is still possible to make assumptions about the magnitude of occurrence that the different compounds have in the sample. The areas in the chromatogram represent the occurrence of that compound in the sample and by that it is possible to determine if a certain compound is highly represented in the sample or not, See Table 9, Appendix A.

From the results of this study, it can be seen that sealants and adhesives were by far the two material samples that emitted the highest amounts of VOC compounds and represents around 75% of all the compounds emitted, see figure 4. The emitting VOCs from sealant are

Preschool C Reference Literature Hazardous

1-Butanol VTT, FS, SS X

2,2,4,6,6-Pentamehtyl heptane PE, FS X

2,6-di-tert-butyl-p-cresol IP X Decanal PE, VTT Dodecane VTT, FS, PE X Hexadecane FS X Hexanal SS, FS Nonanal PE, VTT X Propylene glycol FS Tetradecane FS X Tridecane PE, FS X Undecane VTT

(21)

approximately 90% glycol ethers as can be seen in figure 7, while the adhesive emits mainly alkanes, siloxanes and various unidentified compounds, figure 6. The glycol ethers are

frequently used as solvents for different materials and the presence of these compounds in the sealant sample might be traces from the manufacturing process. Sealants and adhesives were also the main emitters in both number of compounds and in the amount of compounds in a report by Swedish chemicals agency (8). The main compound in the analyzed sealant was propylene glycol which is mainly used as solvents and diluents in, for example,

pharmaceutical preparations, arts and crafts, paint, agricultural products and personal care products. (11) In the air samples from preschool C, propylene glycol was also detected, but the detected amounts are probably more likely caused by the presence of various paints, solvents and other hobby materials that is associated with day care activities. (11)

In the adhesive, silicon based compounds, such as siloxanes and silanes, are present in high amounts. This was, however, not surprising since silicon is widely used in adhesives and various building materials as fillers and corrosion inhibitors. Other compounds of interest identified in the adhesive were the three isomers of xylene (meta, ortho and para) that are dimethylated alkyl benzenes. Because of their similar structures, it was difficult to distinguish which peak that represent which compound, however, some studies have shown that m- and p-xylene elutes at the same time and o-xylene elutes shortly after. (28) P-xylene is known to be used in plastic film and rubber for consumer products, while both the o- and m-xylene are more common in industry usage as intermediates, fuels and fuel additives. Xylenes are also known to be used as solvents in paint, adhesives and flooring. (8, 11) Xylenes were identified in the paint sample but not in the adhesive or linoleum flooring samples. The absence of xylenes in the latter two samples might be due to that the manufacturers used a different solvent for their products. Socialstyrelsen (1) lists xylenes as the most common aromatic hydrocarbon after toluene and that they are used as solvents in, for example, paints, sealants and adhesives, which might explain their presence in the air samples. All three isomers are classified as hazardous and affect, for example, the respiratory system, eyes, skin, central nervous system and blood. (29)

Insulation and paint were low emitters based on the TVOC, see Figure 4. The low VOCs in insulation was not unexpected since its main content is minerals and/or glass fibers, the paint, however, was expected to emit more VOCs because paint in general contains various organic compounds such as 1-butanol, hexanal and undecane. (8, 13) The plastic film samples and linoleum flooring samples had some varieties among the sampled materials from the different preschools and was most likely due to different manufacturers or some minor contamination before sampling. Alkanes was, however, a major emitter in all three plastic film samples and the linoleum flooring samples both emitted aldehydes.

Information from different reports and studies was used to compare results obtained in this study. The sources include the National Board of health and Welfare (Socialstyrelsen) (1), the Swedish chemicals agency (Kemikalieinspektionen) (8) and a study by Helena Järnström (13). Eurofins Pegasuslab AB was used as a reference lab and they analyzed insulation and 3M tape from preschool B, T-Flex tape and plastic film from preschool A and the matched compounds are shown in table 6-8. The Swedish chemicals agency has listed 46 selected dangerous compounds that are known to be used in building materials (8), of those 46

compounds, three have been identified in this study; toluene, n-hexane and 2,6-di-tert-butyl-p-cresol. These compounds are known carcinogens, mutagenics and disturbers of propagation. (8) N-hexane and toluene was mainly identified in the adhesive while the 2,6-di-tert-butyl-p-cresol was identified in the insulation from preschool C. Toluene are also listed as present in adhesives and paint in kemikalieinspektionens report (8) but not in the study by Järnström, H. (13). Toluene and n-hexane have also been identified in the indoor air samples from the preschools.

(22)

1-butanol was the only compound that was identified in air samples from all preschools and also identified in material samples from all three preschools such as sealant, acrylic, T-Flex tape, linoleum flooring. 1-butanol is known as a degradation product when phthalates in adhesives hydrolyze PVC-flooring, other known sources is detergents and polishes for flooring and furniture. (1) 1-butanol is, however, not mentioned at all in

kemikalieinspektionens report (8) but is listed in Helena Järnströms study as common

identified in PVCs, adhesives and paints. (13) 1- butanol was found in the T-flex tape sample, which is interesting since the technical data sheet for T-flex tape, states that the material was free from solvents. 1-butanol is used as a solvent but also as intermediates and fuel. (26) Tridecane, however, was identified in material samples from all three preschools but was only detected in air samples from preschool A and C. Propylene glycol which have been a great contributor to the TVOC was identified in material samples from preschool B and C and also in the air samples from the same preschools. Propylene glycol was neither listed as common in building materials by Järnström, H. (13) or as one of the common hazardous compounds by kemikalieinspektionens report (8).

The reference lab, Eurofins Pegasuslab AB, identified compounds in their analysis that were not identified in this study, for example, 2,2,4,6,6-pentametylheptan was highly present in insulation and plastic film, it was also present in the two tape samples.

2,2,4,6,6-pentametylheptan is used in the manufacturing of thermoplastics and might explain their prescense in the samples, it is however remarkable that the compound was only identified in the insulation samples from preschool B and C. Only a total of six chemical compounds were detected in both the reference lab analysis and the analysis in this study. The reason for the variety in values in the two analyses may be due to several reasons, such as, contamination and variations in the evaluation of the samples.

Emissions of OPFRs

The low levels of detected OPFR compounds in the building materials may be due to the absence or low added amounts of OPFRs in the selected building materials but also low emission rates of the compounds. Since OPFR compounds are semi volatile, they can also adsorb to surfaces once they have been emitted from the material, such as onto the inside glass walls of the test cell. Furthermore, some OPFRs have been detected in the dust samples from the preschools which suggests that dust can also be a major matrix for these compounds. Furthermore, other sources of OPFRs in the indoor environment could also be furniture, electronics, textiles and other. (20) Because of the absence of emission-rates, it is not possible to calculate accurate concentrations in the samples but only to make assumptions whether a sample contains OPFR compounds or not.

TCIPP was the only compound detected in all the building material samples and in dust samples from two preschools. The compound is one of the most commonly identified flame retardant occurring in indoor environments which may be explained by that it is classified as a SVOC, which is known to have re-emission properties. It has also been suggested that the compound is more resistant to abiotic degradations (30, 31) TCIPP is used as a flame retardant but also as a chemical constituent for adhesives, sealants and various building materials which explains its high presence in the material samples. (26) T-Flex tape had six identified OPFR compounds, TnBP, TCEP, TCIPP, TPHP, TBOEP and EHDPP, all are known flame retardants but some of them are also used as plasticizers, solvents and in building products, such as TCIPP, TPHP, EHDPP and TBOEP. This might also explain the high presence of OPFR compounds in the T-Flex tape sample since the material is designed to be extremely sticky, air and damp proof, age resistant and resistant to cold and heat, and therefore consists of various chemical compounds that generates these properties. (26)

(23)

Building materials

When comparing the same types of materials from the different preschools, such as insulation, plastic film, linoleum flooring and tape, they showed similar chemical content, figure 5-7. Small differences were, however, noticed, for example, the dominant chemical group for insulation from preschool B were amines and alkanes while for insulation from preschool C they were aromatic hydrocarbons, various unidentified compounds and alkanes. The main content of linoleum floorings was aldehydes, the linoleum flooring from preschool B,

however, showed high presence of ketones while linoleum flooring from preschool C did not, figure 5-7. The reasons for these variations of chemical content might be due to several reasons, for example, the different manufacturers, that produces the materials differently according to constituents. The age of the materials might also be a source because the various compounds may have emitted already.

Analysis and evaluation

VOCs

The evaluation of the VOC chromatograms from the GC/MS analysis was performed with a threshold value applied to sort out unwanted noise peaks. Basically, a higher threshold gave less peaks because everything below the threshold is treated as noise. The threshold settings in this study are between 19 to 22 depending on the amounts of signals in the chromatogram, See Table 2. The influence of threshold values in the evaluation of chromatogram became

particularly evident for the adhesive, with a threshold of 22 and a number of 74 peaks was identified and with a threshold of 20 approximately 130 peaks were identified. See figure 8, 9 in Appendix B.

The identification of the compounds represented by each peak in the chromatogram was performed by matching the ionization pattern against the NIST-2011 MS Library (M Search 2.0, NIST). Peaks that had an 80% to 100% ionization pattern similarity to a certain

compound was considered a safe assumption while peaks that had a similarity below 80% was considered too uncertain to assume and was labeled as various. The uncertainty to distinguish between various compounds became especially evident for samples that is believed to contain silicon compounds due to the similarity of mass by siloxanes (Si-O-Si) and phenols (C6)

which both has a base weight of 72 u. The similarity of ionization patterns for both siloxanes and phenols made it hard for the NIST library to assign a compound and therefore these compounds were added to the various category. Retention index (RI) is a standardized retention time method where linear n-alkanes is used as standards to obtain a linearity between the carbon atom numbers in 1D GC, and is used to aid in identification of compounds. The usage of RI in this study would most likely add more certainty in the identification of the compounds that had a match in the NIST library below 80%. (33, 34) Contamination from the 20mm butyl injection stoppers (Grey) used to seal the flasks became apparent when the VOC blanks were evaluated. Early samples were sealed with the 20 mm rubber stoppers (Red), but since they were no longer available for the remaining samples and the grey caps was used instead. The contamination from the grey caps was considerable stronger than the red caps, this was visible by the number of peaks each blank analysis contains, see table 2. The grey cap blank chromatogram showed 10 peaks with a total volatile organic compound (TVOC) area ten times as large as the red cap blank. The red cap blank showed a remarkable low peak count of 3 peaks where the largest peak was the internal standard, See figure 10, 11 in Appendix B.

In the plastic film sample from preschool C, a blank sample were added by mistake during fractionation and therefore, the values in the plastic film sample cannot be treated as accurate values since it consists of two samples.

(24)

OPFRs

The internal standard recovery was all within the margin of 50-120%, except for TEP and TnPP that often showed so low values that no recovery could be calculated. This was probably because TEP and TnPP compounds are very volatile and therefore have been lost during the evaporation, the high values, however, may be the product of matrix effects. This problem has also been encountered in Van den Eede et al. (26).

Conclusions

The emissions from building materials used in low energy preschools may contribute to the presence of different hazardous compounds in the preschools indoor environment. However, this study does not clarify to what extent they contribute and it is important to consider other possible sources when evaluating the total indoor environmental quality, such as furniture, outdoor air and various consumer products used.

It is also important to keep in mind that this was a qualitative study, concentrations of the identified compounds have not been calculated. This means that the numerous identified hazardous compounds may exist in the analyzed material samples but it is however not determined if the levels they exist in are hazardous.

Identified compounds of interest in this study were the three isomers of xylene (meta, ortho and para), toluene, n-hexane and propylene glycol, all but the last compound are associated with hazardous properties such as affection of the central nervous system, temporary incapacitation, skin and eye irritants and disrupters of the respiratory system.

A minor amount of 13 hazardous compounds have also been identified in the indoor air samples, this is however a little misleading observation since over 100 different compounds were identified in the building materials but just a few compounds were evaluated in this study, due to the limited time frame of this work.

TCIPP was the only OPFR compound that was identified in all the materials and may be explained by its ability to resist abiotic degradation and their slow emission rates combined with its ability to re-emission. The amount of OPFRs identified in the dust samples from the preschools was probably not caused by the emission from building materials, this was supported by the low values obtained in the analysis of the various building materials. The problem with low recovery of procedure blanks, due to evaporation, made it difficult to analyze the two more volatile compounds TEP and TnPP, other methods for analyzing these two may be evaluated for further reference.

Acknowledgment

I would like to express my deepest gratitude to Josefin Persson for her support,

encouragement, guidance and patience during this project. I would also like to thank my supervisor Thanh Wang for his support and guidance and Helena Arvidsson at AMM for all the help during sampling and analysis.

(25)

References

1. Socialstyrelsen, September 2006, ”Kemiska ämnen i inomhusmiljön”

2. Burström, P.G., 2007. Byggnadsmaterial : uppbyggnad, tillverkning och egenskaper 2. uppl.., Lund: Studentlitteratur.

3. http://www.energimyndigheten.se/energieffektivisering/byggnader/ 2017-05-09 4. https://ec.europa.eu/energy/en/topics/energy-efficiency/buildings/nearly-zero-energy-buildings 2017-05-09

5. Vojta et al., 2017. Screening for halogenated flame retardants in European consumer products, building materials and wastes. Chemosphere, 168, pp.457–466.

6. Bachmann & Myers, 1995. Influences on sick building syndrome symptoms in three buildings. Social Science & Medicine, 40(2), pp.245–251.

7. Kemikalieinspektionen, Rapport Mars 2011, ”Handlingsplan för en giftfri vardag 2011– 2014, Skydda barnen bättre”

8. Kemikalieinspektionen, PM 9/15, ”Kartläggning av farliga ämnen i byggprodukter i Sverige”

9. Kemikalieinspektionen, Rapport nr 8/13 ”Barns exponering för kemiska ämnen i förskolan”

10. Xu, F. et al., 2016. Comprehensive Study of Human External Exposure to

Organophosphate Flame Retardants via Air, Dust, and Hand Wipes: The Importance of Sampling and Assessment Strategy. Environmental science & technology, 50(14), pp.7752– 60.

11. https://pubchem.ncbi.nlm.nih.gov/ 2017-05-22

12. www.nollhus.se, FEBY 12, Jan 2012, ”Kravspecifikation för nollenergihus, passivhus och minienergihus”

13. Järnström, Helena., 2011. Emissioner av kemiska föreningar från byggnadsmaterial. VTT Expert Services

14. Salthammer, T., Bahadir, M. & Schleibinger, Hans, 2009. Occurrence, Dynamics and Reactions of Organic Pollutants in the Indoor Environment. CLEAN – Soil, Air, Water, 37(6), pp.417–435.

15. Afshari, A., Lundgren, B. & Ekberg, L.E., 2003. Comparison of three small chamber test methods for the measurement of VOC emission rates from paint. Indoor Air, 13(2), pp.156– 165.

16. ISO 16000-6:2011

17. VOC emissions testing standard, prEN 16516

18. Weschler & Nazaroff, 2008. Semivolatile organic compounds in indoor environments. Atmospheric Environment, 42(40), pp.9018–9040.

(26)

19. http://ki.se/imm/flyktiga-organiska-amnen-voc 2017-05-25

20. Cristale et al., 2016. Occurrence and sources of brominated and organophosphorus flame retardants in dust from different indoor environments in Barcelona, Spain. Environmental Research, 149, pp.66–76.

21. Mankidy et al., 2014. Effects of novel brominated flame retardants on steroidogenesis in primary porcine testicular cells. Toxicology Letters, 224(1), pp.141–146.

22. Abdallah, M.A.-E. & Covaci, A., 2014. Organophosphate flame retardants in indoor dust from Egypt: implications for human exposure. Environmental science & technology, 48(9), pp.4782–9.

23. Cequier, E. et al., 2014. Occurrence of a broad range of legacy and emerging flame retardants in indoor environments in Norway. Environmental science & technology, 48(12), pp.6827–35.

24. Kademoglou et al., 2017. Legacy and alternative flame retardants in Norwegian and UK indoor environment: Implications of human exposure via dust ingestion. Environment International, 102, pp.48–56.

25. EUROPEAN COMMISSION HEALTH & CONSUMER PROTECTION

DIRECTORATE-GENERAL, “Risk assessment Report on Tris (nonylphenyl) phosphite (TnPP)” 26. https://pubchem.ncbi.nlm.nih.gov/ 2017-05-22 27. ISO 16017-2:2003 28. http://www.restek.com/Technical-Resources/Technical-Library/Environmental/env_EVAN1725B-UNV 2017-05-22 29. http://www.chemspider.com 2017-05-23

30. Van Den Eede et al., 2016. Kinetics of tris (1-chloro-2-propyl) phosphate (TCIPP) metabolism in human liver microsomes and serum. Chemosphere, 144, pp.1299–1305. 31. Liu, Allen & Roache, 2016. Characterization of organophosphorus flame retardants’ sorption on building materials and consumer products. Atmospheric Environment, 140, pp.333–341.

32. Afshari, A., Lundgren, B. & Ekberg, L.E., 2003. Comparison of three small chamber test methods for the measurement of VOC emission rates from paint. Indoor Air, 13(2), pp.156– 165.

33. Hérent, De Bie & Tilquin, 2007. Determination of new retention indices for quick identification of essential oils compounds. Journal of Pharmaceutical and Biomedical Analysis, 43(3), pp.886–892.

34. Wang et al., 2016. A regression model for calculating the second dimension retention index in comprehensive two-dimensional gas chromatography time-of-flight mass

(27)

(28)

Appendix A

Table 9 | Selected compounds identified in chromatograms.

Compound Corr. Area Material Group

6-mehtyl-1-heptanol 44134 3M tape Alcohol

Dodecyl acrylate 4701 3M tape Ester

2-Ethylhexylacrylate 2669 3M tape Ester

1-(2-butoxietoxi)-ethanol 676969 Acrylic Alcohol, Glycolic ether Propylene glycol 117119 Acrylic Alcohol, Glycolic ether

Ethylene glycol 95644 Acrylic Alcohol

1-butanol 66626 Acrylic Alcohol

Diethylene glycol 23593 Acrylic Alcohol, Glycolic ether Trans-2-dodecane-1-ol 13047 Acrylic Alcohol

Undecane 10115 Acrylic Alkane

Dodecane 8042 Acrylic Alkane

Glycerin 8025 Acrylic alkaloid

Nonanal 5684 Acrylic Aldehyde

Decanal 3794 Acrylic Aldehyde

Benzaldehyde 1687 Acrylic Aldehyde

Butanal 1224 Acrylic Aldehyde

1,5-dimetoxy-1,3,5-trimethyltriSiloxanes 1766182 Adhesive Siloxaneser TrimetoxiphenylSilanes 1373684 Adhesive Silanes

TBDMS-derivativtrimetoxibenzylalcohol 392540 Adhesive Aromatic hydrocarbon

Toluene 341807 Adhesive Alkene, Aromatic Hydrocarbon

TrimetoximethylSilanes 295392 Adhesive Silanes TrimetoxivinylSilanes 286908 Adhesive Silanes Chlorotris(p-tolyl)methane 187272 Adhesive Various

2,4,5-trimetoxi-N-methylbenzenmethaneamine 180872 Adhesive Various 2,5-dimethyl-1,4-dioxan 173789 Adhesive Various 3-(trimetoxysilyl)-1-propanamine 156633 Adhesive Various

n-hexane 116922 Adhesive Alkane

DimethoxydimethylSilanes 109689 Adhesive Silanes

N,N-dimethylformamid 88695 Adhesive Amine?

2-hydroxibensylsulfonetyl 83735 Adhesive Phenols

Decane 79621 Adhesive Alkane

m-xylene 61257 Adhesive Alkene, benzene

o-xylene 56595 Adhesive Alkene, benzene

3,4,5-trimethoxi-ethylEster bensene

propanoic acid 55634 Adhesive Various

2(ethylamineoetanol) 54122 Adhesive Varioust

Dodecane 52796 Adhesive Alkane

Ethylacetate 52244 Adhesive Ester

Benzen 41214 Adhesive Alkene, benzene

(29)

p-xylene 31856 Adhesive Alkene, benzene DimethoxiphenylSilanes 28846 Adhesive Siloxanes

1,2-dimethoxy-4(1-methoxy-1-propenyl)benzen 24986 Adhesive Siloxaneser

Octane 20649 Adhesive Alkane

Methylcyclopentane 17873 Adhesive Alkane, Cyclic

Tetradecane 13673 Adhesive Alkane

Cyclohexane 12185 Adhesive Alkane, Cyclic

1-methyl-cyclododecane 30438 Insulation Alkane, Cyclic n-ethyl benzenamine 22215 Insulation Varioust 1-methyl-cyclo-dodecene 22119 Insulation Alkene, cyclic 1-ethyl-benzenamine 12324 Insulation Various 2,6-olitertbuthyl-p-cresol 5039 Insulation Alcohol

Decanal 2764 Insulation Aldehyde

Benzotiazol 2762 Insulation tiazol

Nonanal 2381 Insulation Aldehyde

2,2,4,6,6-pentamethyl heptane 2071 Insulation Alkane 2,2,4,6,6-pentamethyl heptane 1846 Insulation Alkane

Benzotiazol 1809 Insulation Various

Decanal 1721 Insulation Aldehyde

Nonanal 1417 Insulation Aldehyde

Hexanal 219936 Linoleum Aldehyde

1-Penten-3-ol 92128 Linoleum Alcohol

Pentanal 65357 Linoleum Aldehyde

Pentanal 63838 Linoleum Aldehyde

Octanal 45926 Linoleum Aldehyde

1-Pentanol 39505 Linoleum Alcohol

Octanal 38622 Linoleum Aldehyde

Nonanal 37018 Linoleum Aldehyde

Propanoic acid 34286 Linoleum Karboxylic acid

1-Pentanol 27152 Linoleum Alcohol

Heptanal 26743 Linoleum Aldehyde

Nonanal 23598 Linoleum Aldehyde

2-Heptanon 22853 Linoleum Ketone

Heptanal 21214 Linoleum Aldehyde

Heptane 14874 Linoleum Alkane

1-Butanol 14657 Linoleum Alcohol

Hexamethylcyclotrisiloxanes 13936 Linoleum Siloxanes

Cyclohexanon 13585 Linoleum Ketone, Cyclic

Octanoic acid 9810 Linoleum Karboxylic acid

2-Heptanon 9355 Linoleum Ketone

Butanoic acid 8484 Linoleum Karboxylic acid

1-Butanol 7680 Linoleum Alcohol

(30)

1-Okten 7394 Linoleum Alkene

Dekanal 7028 Linoleum Aldehyde

2-Butenal/KrotonAldehyde 6940 Linoleum Aldehyde

1-Decene 6639 Linoleum Alkene

2-Nonanon 6163 Linoleum Ketone

Bensaldehyde 5894 Linoleum Aldehyde, Cyclic

2-Hexanon 5869 Linoleum Ketone

2-Pentanal 5790 Linoleum Aldehyde

Pentylformat 5771 Linoleum Ester

Pentylformat 5461 Linoleum Ester

3-Pentanon 4817 Linoleum Ketone

Bensaldehyde 4047 Linoleum Aldehyde

1-Decene 3760 Linoleum Alkene

Octanoic acid 2775 Linoleum Karboxylic acid

1-Hepten 2449 Linoleum Alkene

Decanal 1851 Linoleum Aldehyde

Dietylenglykol 26002 Paint Alcohol, Glycolic ether

o-xylene 10482 Paint Alkene, benzene

m-xylene 8566 Paint Alkene, benzene

Propylen glycol 7473 Paint Alcohol, Glycolic ether

p-xylene 4246 Paint Alkene, benzene

n-hexan 1615 Paint Alkane

Toluene 1586 Paint Aromatic Hydrocarbon

Nonanal 1470 Paint Aldehyde

Decanal 1120 Paint Aldehyde

Amineoetane 3922 Plastic film Amine

Amineoetane 3922 Plastic film Various

1-Methoxi-2-propanol 2490 Plastic film Alcohol, Glycolic ether 1-Methyl-cyclododecane 2093 Plastic film Alkane, Cyclic

2-Ethylhexylacrylate 1800 Plastic film Ester 2-Ethylhexylacrylate 1800 Plastic film Ester

Nonanal 1577 Plastic film Aldehyde

Nonanal 1281 Plastic film Aldehyde

Decanal 1229 Plastic film Aldehyde

Tridecane 1086 Plastic film Alkane

Hexanal 990 Plastic film Aldehyde

Dodecane 978 Plastic film Alkane

Tetradecane 938 Plastic film Alkane

n-Ethylbezenamine 840 Plastic film Various

Hexadecane 801 Plastic film Alkane

1-Methoxi-2-propanol 775 Plastic film Alcohol, Glycolic ether

(31)

Tetradecane 772 Plastic film Alkane Hexamethylcyclotrisiloxanes 700 Plastic film Siloxanes Hexamethylcyclotrisiloxanes 642 Plastic film Siloxanes

Tridecane 603 Plastic film Alkane

1-Methoxi-2-propanol 564 Plastic film Alcohol, Glycolic ether Propylene glycol 6364437 Sealant Alcohol, Glycolic ether 1-acetat-1,2-propandiol 146584 Sealant Alcohol, Glycolic ether

Alkene C12 123376 Sealant Alkene

1-Butanol 116788 Sealant Alcohol

2-acetat-1,2-propandiol 67488 Sealant Alcohol, Glycolic ether

Alcohol C8-C10 38084 Sealant Alcohol

diethylen glycol 30796 Sealant Alcohol, Glycolic ether

Ketone C10 22874 Sealant Ketone

2-Ethylhexanol 60398 T-Flex tape Alcohol

1-butanol 15813 T-Flex tape Alcohol

2-Ethylhexylacrylate 15766 T-Flex tape Ester

Aminoetane 9664 T-Flex tape Amine

Hexanal 1863 T-Flex tape Aldehyde

(32)

Appendix B

Figur 8 | Chromatogram of Adhesive from preschool B, threshold 22.

(33)

Figur 10 | Chromatogram of the grey cap blank.

(34)

Appendix C

Table 10 | Calculation of MDL from the procedure blanks from the four batches. *one of the blanks for TBOEP was treated as an outlier due to its high value and was not calculated.

Name Mean Standard deviation MDL

𝑥̅=∑ 𝑥 𝑛 𝜎 = √ ∑(𝑥 − 𝑥̅)2 𝑛 3×σ TnBP 1030 447 1341 TCEP 285 20 59 TCIPP 3801 582 1746 TDCIPP 946 541 1623 TPHP 589 134 401 TBOEP* 1195 133 398 EHDPP 508 178 534

Identification of volatile organic compounds (VOC) and organophosphate flame retardants (OPFR) in building materials